How to Merge Songs from YouTube into One MP3 — Properly

Cut each track to the part that matters, set the order, export once. The whole thing takes about two minutes. Four small decisions separate a mix that sounds edited from one that sounds broken.

There's a particular kind of friction that comes from wanting something simple.

You've got six YouTube tabs open. Two live versions, a remix someone uploaded in 2013, a track you want only the back half of. What you actually want is one file — a single MP3 you can drop onto your phone and stop thinking about.

The standard advice is to download each one separately and stitch them together in Audacity. That works. It also means installing a DAW, learning a timeline, and spending twenty minutes on something that has no business taking twenty minutes.

There's a shorter path. The mechanics take two minutes, and then there are four decisions that determine whether the result sounds deliberate or amateur. Most guides skip those entirely, which is why most homemade mixes have that telltale lurch between tracks.

The two-minute version

1. Paste your first YouTube link.
2. Add another source, paste the next. Repeat — up to 25 clips.
3. Set a start and end time on each one, keeping only the part you want.
4. Put them in the order you want to hear them.
5. Export. One MP3, 320 kbps, done.

That's the whole flow for a YouTube to MP3 merger that works in the browser — no install, no account. The rest of this is about doing it well.

First, trim ruthlessly

Before fades, before order, before anything: cut the fat.

A five-minute "song" on YouTube is often four minutes of music wrapped in a channel intro, a sponsor read, and twenty seconds of end cards. On a single download you'd shrug. In a merged file, that's the thing that makes it unlistenable — you've built a mix where every third minute is someone asking you to subscribe.

So scrub to where the music actually starts and where it actually ends. Be unsentimental about it. If your tool doubles as a YouTube to MP3 trimmer that takes exact timestamps rather than a drag handle, use that — typing 1:04 beats eyeballing it every time.

Two judgement calls worth making consciously:

Live recordings. Decide whether you're keeping the applause. Sometimes it's a lovely bridge into the next track. Sometimes it's forty seconds of a crowd in a stadium you weren't at.

Rhythmic mixes. If this is for running or the gym, trim to a musical boundary — the top of a bar, not just the first audible sound. It costs you two seconds of scrubbing and it's the difference between a transition that lands and one that trips.

Then fade — but only a little

Here's the thing nobody tells you: if you cut a song at an arbitrary point, the waveform stops mid-cycle. You get a click. Sometimes a thud. Stack five of those and the file doesn't sound edited, it sounds damaged.

The fix is small. Roughly 0.3 to 1 second of fade in, 1 to 2 seconds out. That's enough to kill the click while staying short enough that nobody perceives it as a fade at all. It just sounds like the track ended.

Longer fades — three seconds and up — read as intentional, and they're genuinely useful between tracks of very different energy. But they eat music. Don't put a four-second fade on a tight cut you worked to get right.

One honest note on terminology, because it matters when you're choosing a tool: joining clips end to end is concatenation, not crossfading. The tracks play one after another; they don't overlap. If you want one song rising as the previous one falls away, that's a different operation and a dedicated crossfade tool will do it better. For a mix where each track gets room to breathe, fades on both ends are what you wanted anyway.

The volume problem nobody diagnoses

This is the one listeners notice most and almost nobody identifies.

Two uploads of the same song can differ by several decibels. A loud remaster sits next to a quiet soundboard recording sits next to something captured on a phone in 2011. While you're browsing YouTube you never notice, because YouTube normalises playback loudness on the fly. The moment you extract the audio, that normalisation is gone — and your mix has a track that blows your headphones off followed by one you can barely hear.

Adjust volume per clip, not on the finished file.

A method that works: play each source, find the one that sounds most normal, and treat it as your reference. Bring everything else to meet it. And pull the loud ones down rather than pushing the quiet ones up — boosting a quiet source lifts its noise floor and tape hiss right along with the music.

Don't chase perfection. Getting within a couple of dB is enough that nobody reaches for the volume knob halfway through.

Order is a creative decision, not a technical one

Once the clips are trimmed and levelled, sequencing is the only interesting part left — and it's pure taste.

The one piece of structural advice I'd offer: don't front-load everything good. A mix that opens at maximum intensity has nowhere to go. Give it an arc, even a crude one.

There is one technical thing happening underneath that's worth understanding, even though you shouldn't have to manage it. To join audio files, every input needs the same sample rate and channel layout. YouTube serves Opus at 48 kHz and AAC at either 44.1 or 48 — so a handful of sources gives you a handful of formats. Concatenate them without normalising first and you get anything from pitch-shifted audio to a file that simply won't play.

Any decent browser-based tool handles this silently — TubeMixLab brings everything to 44.1 kHz stereo before joining. It's worth knowing because it's by far the most common reason a hand-rolled ffmpeg concat produces a broken file, and if you ever do this manually, it's the first thing to check.

Quality settings, briefly

Convert YouTube to MP3 at 320 kbps for music. Always. The source is already lossy — YouTube hands you a compressed stream — so re-encoding at 128 kbps compounds loss on audio that's been through one encoder already. The storage you save isn't worth it on something you'll actually listen to.

For spoken word — lectures, interviews, podcasts — 128 kbps is genuinely fine and the file is a third the size.

Length adds up faster than people expect:

Fine to download. Mildly annoying to email.

What you won't get

I'd rather say this up front than let you discover it.

No track metadata. A merged file has no artist, title or cover art written into it, so it arrives in your music library as an untagged file. That's inherent to the operation — whose artist would it be, when the file is four different ones? — but it means you'll want to give it a sensible filename on the way out, or tag it in your player afterwards.

No crossfade. Covered above, but worth repeating as a limitation rather than a footnote.

Exported files are also deleted from the server after 30 minutes. Download it before you wander off to make tea.

When a different tool is the right answer

Honest version, because the internet is full of guides that pretend there's one answer to everything:

• You want tracks blending into one another → a dedicated audio joiner with crossfade. Concatenation won't do it, no matter how you set the fades.
• Your audio is already on your computer → a file-based editor. Browser tools that take YouTube links generally can't take uploads.
• You need real production — multitrack, EQ, automation, beatmatching → that's a DAW, and nothing in a browser tab substitutes for one.
• You just want one song as an MP3 → any converter does that. You don't need a merger for a single track.

The case this is actually for is narrow and specific: several YouTube videos, each needing its own trim, out as one clean file, without installing anything. If you want to cut and merge songs online from YouTube and that's the thing you came here for, two minutes is genuinely all it takes.

Published 15 October 2026